1242301 九、發明說明: • 【發明所屬之技術領域】 本發明係有關一種高效率省電發光二極體驅動方法,尤指一 種發光二極體被動矩陣藉由列(R〇W)、行(C〇lumn)中之各種相 位(Dis-Charge、Pre-Charge、Reverse Bias、Floating、Current ON/Sink)的相互搭配,來達到最佳化效能的驅動方式。 【先前技術】 按,隨著電子產業的進步,愈來愈多的警示、動畫或廣告方 籲 面產品利用發光二極體(LED)來製作,例如:新式具有動晝的紅 綠:、大型LE1D所組成之廣告看板,皆為發光二極體之應用範圍, 其優點為具有高亮度、壽命長及可製作超大型之看板…等之優 點’因此廣為業者所開發利用。 —如圖一A至圖一D所示,傳統的被動矩陣發光二極體驅動方式 為二段式驅動狀態圖,其係於行(C〇iumn)上區分為放電階段 (Dis-Charge Phase)、預先充電階段(pre—charge Phase)、 電流驅動階段(Current Driving Phase),而在列(R0W)上可 區义為電流下降階段(Current Sink Phase)、反相偏壓階段 φ (Reverse Bias Phase)。 於圖一A中,於行被點亮(c〇lumn) (Active Column)的狀 態依序為: 放電階段(Dis-Charge Phase) 11 :係為消除發光二極體所儲 、 存的電量; 預先充電階段(Pre-Charge Phase) 12 :該預先充電的階段可 • 消除發光二極體之寄生電容,且使下述電流開啟階段可獲得較 佳的起始值; 電流開啟階段(Current ON Phase) 13。 於圖一B中,於不被點亮的行(Non-Active Column)的狀陣 順序為: 5 12423011242301 IX. Description of the invention: [Technical field to which the invention belongs] The present invention relates to a high-efficiency power-saving driving method of a light-emitting diode, especially a light-emitting diode passive matrix through columns (ROW), rows ( Column) The various phases (Dis-Charge, Pre-Charge, Reverse Bias, Floating, Current ON / Sink) are matched with each other to achieve the optimized driving mode. [Previous technology] According to the advancement of the electronics industry, more and more warning, animation or advertising products are produced using light-emitting diodes (LEDs). For example: the new red and green with moving daylight: large The advertising kanbans made up of LE1D are all applications of light-emitting diodes. The advantages are high brightness, long life, and the ability to make very large kanbans. —As shown in Figures 1A to 1D, the traditional passive matrix light-emitting diode driving method is a two-stage driving state diagram, which is divided into Dis-Charge Phase on the line (Coiumn). , Pre-charge phase, current driving phase, and the column (R0W) can be distinguished as the current sink phase, the reverse bias phase φ (Reverse Bias Phase ). In FIG. 1A, the states of the row (clumn) (Active Column) are sequentially as follows: Dis-Charge Phase 11: is to eliminate the power stored and stored by the light-emitting diode; Pre-Charge Phase 12: This pre-charge phase can eliminate the parasitic capacitance of the light-emitting diodes, and make the current starting phase below to obtain a better starting value; Current ON Phase ) 13. In FIG. 1B, the sequence of the non-active column (Non-Active Column) is: 5 1242301
放電階段(Dis-Charge Phase) 14 ; 放電階段(Dis-Charge Phase) 15 ; 放電階段(Dis-ChargePhase) 16 :因為該點不被點亮,故使 該發光二極體之陽極保持持續接地之狀態。 於圖一C中,於列被點亮(Row) (Active Row)的狀態依序 電流下降階段(Current Sink Phase) 17 ; 電流下降階段(Current Sink Phase) 18 ; 電流下降階段(Current Sink Phase) 19 :因該列將被點亮, 故需將發光二極體之陰極接地,以形成導通之順向電流。 於圖一D中,於不被點亮的列(R0W) (Non-Active Row)的 狀態順序為: 反相偏壓階段(Reverse Bias Phase) 20 ; 反相偏壓階段(Reverse Bias Phase) 21 ; 反相偏壓階段(Reverse Bias Phase) 22 :因不被點亮的點不 允許有電流流入,故提供一反向偏壓以阻止電流流入發光二極 體,另一方面有反向電場的存在可以發光二極體獲得較長之壽 命。 φ 請繼續參閱圖二A所示,其中該被動矩陣的架構,其點對點之 間將對彼此寄生電容效應產生影響性,當該被動矩陣所形成之面 板被點亮時,發光二極體之驅動器(Driver)將進入第一個階段 (Phase),該第一個階段為一放電狀態,此時S1〜S4為接地狀態, • 而則為列被點亮之狀態,同時,R2、R3之二列處於列不被點亮的 狀態’且與一反相Vrev電位做連接,請再參閱圖所示,於R2、 • 即之一列並未被掃描,但卻被反相yrev電壓對該二列上所有發光 二極體進行充電,因此該Vrev電位浪費於不動作之發光二極體。 請參閱圖二C所示,當被動矩陣的架構進入第二個階段時,假 0又S1被點冗且S2〜S4不被點亮,S1進入一個不被預先充電階段,而 R1接地且R2、R3連接於Vrev電位,此時矩陣位置於1-1之發光二極 1242301 體由將S1充電至預先充電電位,而導致si之預先充電需要有較大 的電壓或較長的時間,當列數愈多時,寄生電容的效應愈大,而 預先充電電路的負載愈重,同時浪費太多不需支出的電力。 請參閱圖二D所示,當被動矩陣的架構進入電流開啟的階段, 此時被點亮的S1提供電流到發光二極體,S2〜S4、R1接地,R2、R3 則連接Vrev。若VconOVrev時,其C2-1、C3-1可被視為- (Vrev-Vcon)的電位,因充電的幫浦效應將使R2、R3端升高其電 壓準位’並使得Vr2、Vr3>Vrev,此些升高的電位乃為一般經由 電路中的靜電防護二極體(ESD Protecti〇n Di〇de)放電而回復 為電位,而此升高再釋放的電位為一種電能的虛耗。 最後進人掃描電荷階段(Scan Duty),其掃描由R1轉換成R2, 而R卜R3連接Vrev電位,而R2、幻、S2、S3接地,發光二極體中 之電容轉變如圖三A、B、C所示,其中該充電幫浦效能電容電荷將 由圖二A->圖三B->圖三C,此時Vrev電位將重新對電容充電而造成 電能的虛耗。 因此該傳統發光二極體之被動矩陣具有極大改進之空間。 【發明内容】 基於解決以上所述習知技藝的缺失,本發明為高效率省電發 極體,動方法,本發明之主要目的在於針對被動矩陣 cled) ^ / 率且省電的驅動方式,其方法 lei !〇 (C〇1Umn) (Dis-Charge>Dis-Charge Phase 14; Dis-Charge Phase 15; Dis-ChargePhase 16: Because this point is not lit, the anode of the light-emitting diode is kept continuously grounded. status. In FIG. 1C, the state of the row is lighted (Row) (Current Row Phase 17); current sink phase (Current Sink Phase) 18; current sink phase (Current Sink Phase) 19: Because the column will be lit, the cathode of the light-emitting diode needs to be grounded to form a forward current that conducts. In FIG. 1D, the state sequence of the non-lit row (R0W) (Non-Active Row) is: Reverse Bias Phase 20; Reverse Bias Phase 21 Reverse Bias Phase 22: Because no light is allowed to flow into the unlit point, a reverse bias is provided to prevent the current from flowing into the light-emitting diode. On the other hand, there is a reverse electric field. The existence of a light-emitting diode allows a longer lifetime. φ Please continue to refer to FIG. 2A, in which the structure of the passive matrix, its point-to-point will affect each other's parasitic capacitance effect. When the panel formed by the passive matrix is lit, the driver of the light-emitting diode (Driver) will enter the first phase (Phase), the first phase is a discharge state, at this time S1 ~ S4 are grounded, and then the column is lit, at the same time, R2, R3 bis The column is in a state where the column is not lit 'and is connected to an inverting Vrev potential. Please refer to the figure again. At R2, • one column is not scanned, but is reversed by the yrev voltage to the two columns. All the light-emitting diodes are charged, so the Vrev potential is wasted on the light-emitting diodes that do not operate. Please refer to Figure 2C. When the architecture of the passive matrix enters the second stage, false 0 and S1 are redundant and S2 ~ S4 are not illuminated, S1 enters a stage that is not precharged, and R1 is grounded and R2 And R3 are connected to the Vrev potential. At this time, the light emitting diode 1242301 of the matrix position is 1-1. The S1 is charged to the precharge potential, and the precharge of si requires a larger voltage or a longer time. The greater the number, the greater the effect of parasitic capacitance, the heavier the load of the pre-charging circuit, and at the same time wasting too much power without expenditure. Please refer to FIG. 2D. When the structure of the passive matrix enters the current-on stage, the lighted S1 provides current to the light-emitting diode, S2 ~ S4 and R1 are grounded, and R2 and R3 are connected to Vrev. If VconOVrev, its C2-1, C3-1 can be regarded as-(Vrev-Vcon) potential, due to the pumping effect of charging will cause R2, R3 terminals to increase their voltage levels' and make Vr2, Vr3 > Vrev, these raised potentials are generally restored to electric potentials through discharge of electrostatic protection diodes (ESD Protectors) in the circuit, and the raised and released potentials are a waste of electrical energy. Finally, it enters the scan charge stage (Scan Duty), whose scan is converted from R1 to R2, and R3 and R3 are connected to the Vrev potential, and R2, phantom, S2, and S3 are grounded. The capacitance transition in the light-emitting diode is shown in Figure 3A. As shown in B and C, the capacitor charge of the charging pump efficiency will be shown in Figure 2A- > Figure 3B- > Figure 3C. At this time, the Vrev potential will recharge the capacitor and cause power loss. Therefore, the passive matrix of the traditional light-emitting diode has great room for improvement. [Summary of the Invention] Based on solving the above-mentioned shortcomings in the conventional art, the present invention is a high-efficiency power-saving generator and method. The main purpose of the present invention is to drive the passive matrix cled ^ / rate and save power, The method lei! 〇 (C〇1Umn) (Dis-Charge >
βί3δ ' F1〇ating ' Current 0N/SinW 互釔配,來達到隶佳化效能的驅動方式。 為達上述之目的’本發明為—種高效 、订不被點党、列被點亮及列 做為不_動方式輕,其各㈣之動作程賴i括來 1242301 (A) 行被點壳狀悲,其依序執行:放電階段、預先充電 階段、電流開啟階段及放電階段; (B) 行不被點焭狀悲,其依序執行:放電階段、浮動階 段、放電階段及放電階段; (C) 列被點売狀態,其依序執行:電流下降階段、電流 下降階段、電流下降階段、電流下降階段;以及 (D) 列不被點壳狀悲,其依序執行:浮動階段、浮動階 段、反向偏壓階段、浮動階段。 較佳者,該四種狀態更係包含有四個階段,其中於(A)、(B)、 (D)的階段為一較短時間,而(c)階段為'一較長時間。 明治^ 一步ΐ本發明有更私的說明,乃藉由以下圖示、圖號說明及發 月砰、、、田呪明,冀能對貴審查委員於審查工作有所助益。 【實施方式】 以刹^配^^之圖式說明本發明之詳細結構,及其連結關係, 以利於貴審委做一瞭解。 ,參關四所示,其係為本發明提升傳統被動矩陣發光二極 動效能之方法,而本發狀發光二極體主要為朗—冷光板 (Electr〇 Luminescence,EU來做為實施例,當然本發明不限 之,,舉凡所有發光二極體⑽)或有機發光二極體 (0LED) ’ ^為本發明可適用之範圍。本發明之主要技術手段 為將傳統習知三段式驅射式改成為四段式驅財式,βί3δ 'F1〇ating' Current 0N / SinW is a driving method for mutual yttrium matching to achieve optimized performance. In order to achieve the above-mentioned object, the present invention is an efficient, order-free party, columns are illuminated, and columns are light-moving. Shell-shaped tragedy, which is performed in sequence: discharge phase, pre-charge phase, current on phase, and discharge phase; (B) Rows are not truncated, which are performed in sequence: discharge phase, floating phase, discharge phase, and discharge phase ; (C) The column is clicked, and it is executed in order: the current drop phase, the current drop phase, the current drop phase, and the current drop phase; and (D) The column is not clicked and shell-shaped, and it is executed in sequence: the floating phase. , Floating phase, reverse bias phase, floating phase. Preferably, the four states include four stages, among which the stages in (A), (B), and (D) are a short time, and the stage (c) is a 'long time. Meiji ^ One step: The present invention has a more private explanation. The following illustrations, drawing numbers and descriptions are provided for the purpose of helping the review committee in reviewing. [Embodiment] The detailed structure of the present invention and its connection relationship will be described with drawings of brakes and ^^ s, so as to facilitate your understanding of the review committee. As shown in Part 4, it is the method of improving the passive passive matrix light emitting diode's dynamic performance of the present invention, and the hair-emitting light emitting diode is mainly a Lang-Luminescence (EU) as an example. Of course, the present invention is not limited thereto. For example, all the light-emitting diodes (i) or organic light-emitting diodes (OLEDs) are applicable to the present invention. The main technical means of the present invention is to change the conventional conventional three-stage drive-fire type into a four-stage drive type.
Phase) ? 本發明針對:列被點亮(Act i ve Row )、列不被點亮(N〇n_Ac Row) (Active Column) (Non-ActivePhase)? The present invention is directed to: Act i ve Row, Non-Ac Row (Active Column) (Non-Active
Column) 4四種狀態,來做為不同驅動方式調整,而該四 更,包含气四個階段,其中第⑷、⑻、(D)的階段為^ 短時間,而第⑹f皆段為—較長時間,其各狀態之執行動作^下^ 8 1242301 (A) 行被點亮狀態,其依序執行:放電階段51、預先充電階 段52、電流開啟階段53及放電階段54 ; (B) 行不被點亮狀態,其依序執行:放電階段55、浮動階段 56、放電階段57及放電階段58 ; (c)列被點亮狀態,其依序執行:電流下降階段59、電流下 降階段60、電流下降階段61、電流下降階段62 ;以及 (D)列不被點焭狀悲,其依序執行:浮動階段⑽、浮動階段 64、反向偏壓階段65、浮動階段66。 請參閱圖五A所示,於第一個階段「列被點亮」中,其行的動 作方式係與習知技術相同,而依序執行:放電階段51、預先充電 階段52、電流開啟階段53及放電階段54。但不同的是第二階段「列 不被點亮」的階段被替代為浮動階段,該浮動階段之作用為使一 Vrev電位不對任何發光二極體充電,以省下不必要的虛 ,請參閱圖五B所示,接下來進入第二麵段,而依;/執行:放 電階段55、浮動階段56、放電階段57及放電階段58。此時發光二 極體所形成被動矩陣之S1設為行被點亮的狀態,S2、幻設^為^ 不被點亮的狀態,此時R1為動作且敗、肋為不動作狀態。由於S2、 S3、S4、R2、R3皆為浮動狀態,即代表Vpre電位不對任何發光二 • 極體充電,所以vPre只針對C1-1充電,換言之,沒有任何^干擾 預先充電效能的因素,且無浪費的電力。 請參閱圖五C所示,上述預先充電階段被結束後,進入一電流 開啟階段(第三階段),而依序執行··電流下降階段、電流下降 階段、電流下降階段、電流下降階段…四個重覆執行階段。此時 S1 (第一行)(縱向)接上了一個電流源,且S2 (第二行)、 ‘ (第三行)、S4 (第四行)因不被點亮而被接地,R1 (第一列) (橫向)維持電流下降,R2 (第二列)、R3 (第三列)接到一反 相偏壓。請參閱圖五D及圖五E所示,由於C2-;l、C2-3在R2與R3端 被接上了Vrev電位,而使得幻端的電壓由於電容之充電幫浦效應 (Charge Pump Effect)而上升,造成了電流(Current)的上$ 1242301 時間縮短,而可使發光二極體之發光效能增加,因為訐⑼與仏如 數值相近,故如圖五E中,該Vrev對電容充電(Vrev-Vcon)的電壓 i而較習知方法中該電容充飽Vrev電壓後又完全被釋放電壓 來=省電。故C2-1、C3-1由於電流開啟及vrev電壓同時動作,將 使得反向偏壓階段與電流開啟更為省電。Column) 4 four states to adjust for different driving methods, and the four more, including the four phases of Qi, of which the ⑷, ⑻, (D) stage is ^ short time, and the ⑹f are all- For a long time, the execution of each state ^ down ^ 8 1242301 (A) The line is lit, and it is executed in order: discharge phase 51, pre-charge phase 52, current on phase 53 and discharge phase 54; (B) line In the unlit state, they are executed in order: discharge stage 55, floating stage 56, discharge stage 57 and discharge stage 58; (c) When the column is lit, they are executed in order: current drop stage 59, current drop stage 60 , Current drop stage 61, current drop stage 62; and column (D) is not sorrowful, and is executed in sequence: floating stage ⑽, floating stage 64, reverse bias stage 65, and floating stage 66. Please refer to FIG. 5A. In the first stage "columns are lit", the row action method is the same as the conventional technique, and is performed in order: discharge stage 51, pre-charge stage 52, and current on stage. 53 和 更新 步骤 54。 53 and the discharge phase 54. But the difference is that the second stage "the column is not lit" is replaced by a floating stage. The role of this floating stage is to prevent a Vrev potential from charging any light-emitting diodes to save unnecessary imagination. See also As shown in FIG. 5B, the next step is to enter the second stage, and according to / execution: discharge phase 55, floating phase 56, discharge phase 57 and discharge phase 58. At this time, S1 of the passive matrix formed by the light-emitting diode is set to the lighted state, and S2 is set to ^ not to be lighted. At this time, R1 is activated and defeated, and the ribs are not activated. Because S2, S3, S4, R2, and R3 are all floating, which means that the Vpre potential does not charge any light-emitting diodes, vPre only charges for C1-1, in other words, there is no factor that interferes with the precharge performance, and No wasted electricity. Please refer to FIG. 5C, after the above pre-charging phase is ended, it enters a current turn-on phase (third phase), and sequentially executes the current drop phase, current drop phase, current drop phase, current drop phase ... Repeated execution phases. At this time, S1 (the first row) (vertical) is connected to a current source, and S2 (the second row), '(the third row), and S4 (the fourth row) are grounded because they are not lit, and R1 ( The first column) (horizontal) sustain current drops, and R2 (second column) and R3 (third column) are connected to a reverse bias voltage. Please refer to Figure 5D and Figure 5E. Because C2-; l, C2-3 are connected to the Vrev potential at R2 and R3, the voltage at the phantom terminal is due to the charge pump effect of the capacitor (Charge Pump Effect). The increase caused a reduction of the current time of $ 1242301, which can increase the luminous efficacy of the light-emitting diode. Because 讦 ⑼ and 仏 are similar in value, as shown in Figure 5E, the Vrev charges the capacitor ( Vrev-Vcon) voltage i. Compared with the conventional method, the capacitor is fully charged with Vrev voltage and then completely released voltage to save power. Therefore, C2-1 and C3-1 operate simultaneously when the current is turned on and the vrev voltage, which will make the reverse bias phase and current turn on more power-efficient.
請參閱圖五F所示,其中發光二極體會被正常地點亮,且進入 放電階段(第四階段),而依序執行:浮動階段、浮動階段、反 向,壓階段、浮動。此時SpS4、R1處於接地狀態,R2、奶係 為f動狀態,C2-1、C2-2之二列不會因為si端由Vcon接地後所產 生壓降,而必須再重新充飽Vrev電壓,再次印證此一架 被動矩陣驅動方法來得省電。卞稱軚 由上述圖四至圖五F之揭示,即可瞭解本發明確可達到針對被 ,矩陣(Passive Matrix,PM)所形成發光二極體(LED)之點矩 陣電容量分配效應分析,關整出最有效率且省電的鶴方式, ,方法係為藉由列(Row) 、S(c〇lumn)中之各種相位 CDis-Charge ^ Pre^Charge ^ Reverse Bias ^ Floating > Current in )的相互搭配,來達到最佳化效能的驅動方式。故提出本 案之專利申請,以尋求獲得專利權之保護。 綜上所述,本發明之結構特徵及各實施例皆已詳細揭示,而 可充分顯不出本發明案在目的及功效上均深富實施之進步性,極 具產業之利用價值,且為目前市面上前所未見之運用,依專利法 之精神所述,本發明案完全符合發明專利之要件。 唯以J^述者,僅為本發明之較佳實施例而已,當不能以之限定本發 ,所實施之範圍,即大凡依本發明申請專利範圍所作之均等變化盘: 膊利涵__ ’謹請貴絲委^騰,'並祈惠 准,是所至禱。 【圖式簡單說明】 圖一 A係為傳統的被動矩陣發光二極體驅動方式之第一狀態方塊 1242301 圖 第二狀態方塊 圖—B係為傳統的被動矩陣發光二極動方 圖。 β圖,。:、' ο " 式之 第三狀態方塊 β 圖 〇 ' " ',、气 第四狀態方塊 圖 圖 圖二Α係為傳統的被動矩陣發 =:2統的被動㈣發光二極=二= Γ!=ί統的被動矩陣發光二極體之第:架=: 圖三_傳統的發光二極。 圖四係為树_機轉私二^ 圖。 蹬驅動方式之四個狀態方塊 圖五八係為本發明的被動矩陣發光 圖五b係為本發明的被動矩陣發光二極=二;=。 圖五。係為本發明的被動矩陣發光二極二。 圖五E係為本發明的發光二轉變狀態圖。 岡電各第二轉變狀態圖。 圖五F係為本發_被動轉發光二極體之細架構圖。 圖號說明: 11〜放電階段 12〜預先充電階段 13〜電流開啟階段 14〜放電階段 15〜放電階段 1242301 16〜放電階段 17〜電流下降階段 18〜電流下降階段 19〜電流下降階段 20〜反相偏壓階段 21〜反相偏壓階段 22〜反相偏壓階段 51〜放電階段 52〜預先充電階段 53〜電流開啟階段 54〜放電階段 55〜放電階段 56- -浮動階段 57〜放電階段 58- 〜放電階段 59- -電流下降階段 60〜電流下降階段 61- -電流下降階段 62〜電流下降階段 63- -浮動階段 64, -浮動階段 65, -反相偏壓階段 66’ -浮動階段Please refer to FIG. 5F, in which the light-emitting diode is normally lit, and enters the discharge phase (fourth phase), and is performed in order: floating phase, floating phase, reverse, pressure phase, floating. At this time, SpS4 and R1 are in the grounded state, R2 and milk are in the f-moved state. The two columns of C2-1 and C2-2 will not be full of Vrev voltage because of the voltage drop caused by the si terminal being grounded by Vcon. This proves once again that this passive matrix drive method saves power. It is said that from the above disclosure of Figures 4 to 5F, it can be understood that the present invention can indeed achieve the dot matrix capacitance distribution effect analysis of light emitting diodes (LEDs) formed by passive matrix (PM). The most efficient and power-saving crane method is set up. The method is to use the various phases Crow-Charge ^ Pre ^ Charge ^ Reverse Bias ^ Floating > Current in) Drive way to achieve optimal performance. Therefore, a patent application in this case was filed to seek protection from patent rights. In summary, the structural features and various embodiments of the present invention have been disclosed in detail, and it can fully show that the present invention is deeply implemented in terms of purpose and efficacy, and has great industrial use value. According to the spirit of the Patent Law, the present application is completely in line with the requirements of an invention patent. Only those described by J ^ are only the preferred embodiments of the present invention. When the present invention cannot be limited by this, the scope of implementation, that is, the equivalent change plate made in accordance with the scope of the patent application of the present invention: 'Please ask Guisi to appoint,' and pray for the right to pray. [Schematic description] Figure 1 A is the first state block 1242301 of the traditional passive matrix light-emitting diode driving method. Figure B is the traditional passive matrix light-emitting diode dipole diagram. β diagram ,. :, 'Ο " The third state block β diagram of the formula 〇' " ', the fourth state block diagram of the gas Figure 2 A series of traditional passive matrix emission =: 2 system of passive ㈣ emitting diode = two = Γ! = The first of the passive matrix light-emitting diodes: frame =: Figure III_Traditional light-emitting diodes. Figure 4 is a tree_machine-to-private chart. The four state blocks of the pedal driving mode. Figure 58 is the passive matrix light-emitting diode of the present invention. Figure 5b is the passive matrix light-emitting diode of the present invention = two; =. Figure five. It is a passive matrix light-emitting diode two of the present invention. FIG. 5E is a diagram of the two transition states of light emission of the present invention. Gangdian's second transition state diagram. Figure 5F is a detailed structure diagram of the passive-emitting LED. Drawing number description: 11 ~ discharge stage 12 ~ pre-charge stage 13 ~ current on stage 14 ~ discharge stage 15 ~ discharge stage 1242301 16 ~ discharge stage 17 ~ current drop stage 18 ~ current drop stage 19 ~ current drop stage 20 ~ reverse phase Bias stage 21 ~ reverse bias stage 22 ~ reverse bias stage 51 ~ discharge stage 52 ~ precharge stage 53 ~ current on stage 54 ~ discharge stage 55 ~ discharge stage 56--floating stage 57 ~ discharge stage 58- ~ Discharge phase 59--Current drop phase 60 ~ Current drop phase 61--Current drop phase 62 ~ Current drop phase 63--Float phase 64,-Float phase 65,-Reverse bias phase 66 '-Float phase